Infectious Disease Death Among Infants in the U.S
The overall average annual ID IMR for 2008–2009 was 47.5 deaths per 100,000 live births, nearly half of that reported for 1983–1987 (89 deaths per 100,000 live births); however, the proportion of infant deaths due to ID remained similar (8% in 2008–2009 compared with 9% in 1983–1987). The present study found higher ID IMRs in neonates, male infants, LBW and preterm infants (gestational age <37), infants with a 5-minute Apgar score <7, infants of black or AI/AN race and non-Hispanic infants. A maternal age of ≤19 or 20–24 years, unmarried maternal marital status, a live birth order of first or fourth or more, and 0, 1–2 or 3–5 prenatal visits also had higher ID IMRs. Noncongenital ID deaths from 1983 to 1987 also showed higher infant death rates for male infants, LBW infants, infants of black race, infants born to young mothers, infants born to unmarried mothers and infants whose mothers had few or no prenatal visits.
Infants with an ID UCOD are dying early and many of these deaths are potentially preventable; a majority of these deaths are occurring in the first month of life and a large proportion of infants are dying from bacterial sepsis, specifically from an unspecified bacteria. Prevention and intervention should focus on neonatal ID deaths with emphasis on neonatal sepsis. Attention should be paid to appropriate screening for infections in pregnant women. The Centers for Disease Control and Prevention (CDC) along with the American College of OB/GYN and American Academy of Pediatrics recommend that all pregnant women be screened for colonization with Group B streptococcus (GBS) and that colonized women receive intrapartum antibiotic prophylaxis to prevent vertical transmission of GBS to the newborn. This national strategy has proven effective in substantially decreasing the incidence of early-onset neonatal bacterial sepsis due to GBS. However, GBS is still the bacteria most frequently isolated from early-onset neonatal sepsis cultures, particularly among term infants. Missed opportunities for prevention still exist, particularly among women with no or limited prenatal care who are not screened for GBS colonization, women who deliver preterm and those with suspected penicillin allergy, underscoring the importance of strictly following national prevention guidelines. Availability of accurate point of care diagnostics to identify women with colonization at delivery would help identify women without prior screening or those who have become colonized late in pregnancy. In the future, both early- and late-onset neonatal bacterial sepsis due to GBS could potentially be prevented by a safe and effective maternal GBS vaccine.Esherichia coli is another important early-onset neonatal sepsis pathogen, especially among preterm and LBW infants, but there are no recommended strategies for preventing bacterial sepsis due to E. coli.
Racial disparities persist in infant ID mortality. In the present study, infants of black race had a higher ID IMR than that for infants of white race, and black race was associated with increased odds for ID death among LBW infants. Studies of infants in California and North Carolina also found that the mortality rate due to overall ID was highest in infants born to black mothers. In the present study, AI/AN race had a higher ID IMR than white race and was a risk factor in univariate analysis for NBW infants. However, there was no association between AI/AN race and ID death when controlling for other maternal and infant characteristics in both LBW and NBW infants; this finding may be due to the small number of AI/AN infant ID deaths and needs to be confirmed by future studies. It is also of interest to note that Hispanic origin is protective against NBW infant ID deaths in multivariable analysis.
There are some limitations to this study. Some maternal characteristics of interest, such as maternal smoking status, maternal education and adequacy of prenatal care, are potential predictors for infant ID death, but were not comparable between the 1989 and 2003 US Standard Certificate of Live Birth revisions and were not included in the analysis. In addition, there were only a small number of AI/AN infant ID deaths in 2008 and 2009 which could explain why there is a disparity in the ID IMR for AI/AN race but it is not a risk factor when adjusting for the other maternal and infant characteristics in the multivariable analysis. The small number of AI/AN ID deaths may be an effect of racial misclassification; however, we used maternal race and ethnicity as indicated on the birth certificate which are more reliable than infant race and ethnicity reported on the death certificate so this should limit racial misclassification. Also, ICD-10 coding is subject to miscoding or misdiagnosis that could affect inclusion or exclusion of ID deaths. The present study analyzed deaths with IDs as the UCOD which is a conservative approach to identification of ID deaths.
Families and healthcare providers of infants should be aware of the characteristics associated with higher risk of infant ID death. Awareness of these associations and development of further strategic measures should lead to prevention strategies and the reduction of ID deaths among infants.
Discussion
The overall average annual ID IMR for 2008–2009 was 47.5 deaths per 100,000 live births, nearly half of that reported for 1983–1987 (89 deaths per 100,000 live births); however, the proportion of infant deaths due to ID remained similar (8% in 2008–2009 compared with 9% in 1983–1987). The present study found higher ID IMRs in neonates, male infants, LBW and preterm infants (gestational age <37), infants with a 5-minute Apgar score <7, infants of black or AI/AN race and non-Hispanic infants. A maternal age of ≤19 or 20–24 years, unmarried maternal marital status, a live birth order of first or fourth or more, and 0, 1–2 or 3–5 prenatal visits also had higher ID IMRs. Noncongenital ID deaths from 1983 to 1987 also showed higher infant death rates for male infants, LBW infants, infants of black race, infants born to young mothers, infants born to unmarried mothers and infants whose mothers had few or no prenatal visits.
Infants with an ID UCOD are dying early and many of these deaths are potentially preventable; a majority of these deaths are occurring in the first month of life and a large proportion of infants are dying from bacterial sepsis, specifically from an unspecified bacteria. Prevention and intervention should focus on neonatal ID deaths with emphasis on neonatal sepsis. Attention should be paid to appropriate screening for infections in pregnant women. The Centers for Disease Control and Prevention (CDC) along with the American College of OB/GYN and American Academy of Pediatrics recommend that all pregnant women be screened for colonization with Group B streptococcus (GBS) and that colonized women receive intrapartum antibiotic prophylaxis to prevent vertical transmission of GBS to the newborn. This national strategy has proven effective in substantially decreasing the incidence of early-onset neonatal bacterial sepsis due to GBS. However, GBS is still the bacteria most frequently isolated from early-onset neonatal sepsis cultures, particularly among term infants. Missed opportunities for prevention still exist, particularly among women with no or limited prenatal care who are not screened for GBS colonization, women who deliver preterm and those with suspected penicillin allergy, underscoring the importance of strictly following national prevention guidelines. Availability of accurate point of care diagnostics to identify women with colonization at delivery would help identify women without prior screening or those who have become colonized late in pregnancy. In the future, both early- and late-onset neonatal bacterial sepsis due to GBS could potentially be prevented by a safe and effective maternal GBS vaccine.Esherichia coli is another important early-onset neonatal sepsis pathogen, especially among preterm and LBW infants, but there are no recommended strategies for preventing bacterial sepsis due to E. coli.
Racial disparities persist in infant ID mortality. In the present study, infants of black race had a higher ID IMR than that for infants of white race, and black race was associated with increased odds for ID death among LBW infants. Studies of infants in California and North Carolina also found that the mortality rate due to overall ID was highest in infants born to black mothers. In the present study, AI/AN race had a higher ID IMR than white race and was a risk factor in univariate analysis for NBW infants. However, there was no association between AI/AN race and ID death when controlling for other maternal and infant characteristics in both LBW and NBW infants; this finding may be due to the small number of AI/AN infant ID deaths and needs to be confirmed by future studies. It is also of interest to note that Hispanic origin is protective against NBW infant ID deaths in multivariable analysis.
There are some limitations to this study. Some maternal characteristics of interest, such as maternal smoking status, maternal education and adequacy of prenatal care, are potential predictors for infant ID death, but were not comparable between the 1989 and 2003 US Standard Certificate of Live Birth revisions and were not included in the analysis. In addition, there were only a small number of AI/AN infant ID deaths in 2008 and 2009 which could explain why there is a disparity in the ID IMR for AI/AN race but it is not a risk factor when adjusting for the other maternal and infant characteristics in the multivariable analysis. The small number of AI/AN ID deaths may be an effect of racial misclassification; however, we used maternal race and ethnicity as indicated on the birth certificate which are more reliable than infant race and ethnicity reported on the death certificate so this should limit racial misclassification. Also, ICD-10 coding is subject to miscoding or misdiagnosis that could affect inclusion or exclusion of ID deaths. The present study analyzed deaths with IDs as the UCOD which is a conservative approach to identification of ID deaths.
Families and healthcare providers of infants should be aware of the characteristics associated with higher risk of infant ID death. Awareness of these associations and development of further strategic measures should lead to prevention strategies and the reduction of ID deaths among infants.